Horizon CDT Research Highlights

Research Highlights

Expert Moderation of HGV Driving Data: Fair, Trustworthy and Explainable Assessment

  Jimiama Mafeni Mase (2018 cohort)   www.linkedin.com/in/jimiama-mase


The aim of this PhD is to utilise domain experts, computational intelligence and machine learning methods to automatically fuse and regulate multiple sources of data; for a fair and explainable assessment of HGV driving behaviours. The ultimate goal is to develop an information fusion methodology that shows how risky a driver's behaviour is, using multiple sources of data, which are moderated by information obtained from domain experts. The methodology will provide a clear explanation as to which contextual factors influenced the driver's performance. For example, the methodology would determine how risky drivers' behaviours are by fusing their level of distraction, affective state and risk-taking behaviours, and taking into consideration the effects of external factors determined by domain experts such as weather conditions, road geometry, and time pressure for delivery. 


Heavy Goods Vehicles (HGVs) are at the forefront of trade and commerce in the United Kingdom. Both private and public sectors rely on HGVs road transport for the delivery of goods and services. For instance, 1.4 billion tonnes of freight were transported by road between 2016 and 2017 in the UK. Over the same period, a total of 7.8 million tonnes of freight was moved to or from the UK by HGVs. As a result of the importance of HGVs to a nation’s economy, there are great efforts being employed to identify and understand HGV drivers' behaviours that lead to road accidents or incidents. However, most of the efforts do not consider contextual characteristics of driving, such as drivers' physical and mental states, in-vehicle actions (e.g. operation of in-vehicle technologies), weather conditions, traffic conditions, road geometry, work schedules, drivers' reactions to events, other vehicles, and road users. These factors impact drivers' responses and are mostly not captured in the data.

Academic Contributions

  1. Information fusion:
    • ​​​​​​​Explainable fusion of multiple decisions
    • Capture uncertainties in decisions
  2. Driving behaviour:
    • ​​​​​​​Regulate decisions from intelligent driver systems
    • Embed the effects of contextual factors into intelligent driver systems 
    • Improve accuracy and interpretability of machine learning methods for detecting driving behaviours


This author is supported by the Horizon Centre for Doctoral Training at the University of Nottingham (RCUK Grant No. EP/L015463/1) and Microlise.